Bias, Validity偏差，有效性.pptVIP

Medical Research: Validity, Bias, Confounds, Probability, P-value, and Power Kimberly Pendell Validity Validity implies that the measurement reliably measures what it intends to measure. Validity relates to the magnitude of bias. Bias Some different types of bias Selection bias Interviewer bias Recall bias Detection bias Verification bias Publication bias Bias and Research Designs Confounds A confound is a variable that is distributed differently in the study group and the control group and that affects the outcome being assessed. (Riegelman, 2005) Confounding variables can appear through bias or random chance. Probability Probability is the quantitative estimate of the likelihood of condition existing (as in a diagnosis) or of subsequent events (such as in a treatment study). (Guyatt, Rennie, 2002) P-value P-value is the probability of an outcome occurring by chance. The p-value establishes statistical significance. In medical research the standard p-value is p.05 Power Power is the ability of a research study to demonstrate statistical significance. Researchers determine the size of study population needed for a set amount of power before beginning their study. If the study has statistically significant results, the study has enough power. In medical research the standard for power is 80%. Readings: Guyatt, G., Rennie, D. (Eds.). (2002). Users’ guide to the medical literature: a manual for evidence-based practice. USA: American Medical Association. Katzer, J., Cook, K.H., Crouch, W.W. (1998). Evaluating information: a guide for users of social science research. Massachusetts: McGraw Hill. Riegelman, R.K. (2005). Studying a study testing a test: how to read the medical evidence. Philadelphia, PA: Lippincott Williams Wilkins. * * Bias is the systematic tendency to produce an outcome that differs from the underlying truth. Look for unpublished data Publication bias Meta-analysis Blind subjects to research hypothesis Recall bias Case-control study Ran